Loading Experience
0% View on desktop for better experience

KUNAL ARORA

DEVELOPER / AI RESEARCHER

// About

WHO I AM

AI & ML Engineer building intelligent systems that solve real-world problems. B.Tech in Computer Science (AI & ML) at UPES Dehradun.

From lifelike AI companions to LLM-powered interview platforms and edge-deployed vision systems, I operate at the intersection of research and production.

~1 Years Exp
10+ Projects
6 Certifications
// Experience

WHERE I'VE WORKED

Rabbitt AIDelhi Jan 2026 - Present
AI Researcher
  • Led end-to-end development of an AI-driven marketing intelligence platform analyzing company AI visibility through GEO/AIEO signals
  • Designed product architecture from scratch, defining research direction, system design, and scalable backend infrastructure
  • Developed AI pipelines to evaluate brand presence across LLM-driven discovery platforms and generative search interfaces
  • Optimized backend systems and model-serving workflows, reducing server load and improving inference efficiency
// Experience

WHERE I'VE WORKED

Foton LabsLos Angeles (Remote) Oct - Nov 2025
AI & ML Engineer (Contract)
  • Worked on Project GRACE, a lifelike interactive AI companion with real-time 3D visuals
  • Optimized prompt architecture to enhance response relevance and contextual coherence
  • Automated backend server workflows with deployment scripts and health checks
  • Reduced cold-start time by optimizing model initialization across edge nodes
  • Minimized latency using edge functions and distributed execution for near real-time performance
// Experience

WHERE I'VE WORKED

Binary KeedaDehradun Aug - Dec 2025
AI & ML Intern
  • Built LLM-powered interview system with auto-generated personalized questions
  • Developed rubric-based grading engine with JSON scores and feedback
  • Integrated integrity checks: eye-tracking (OpenCV) and AI-plagiarism detection
  • Optimized prompt design and model selection (GPT-4, Claude, Llama-3) for quality and cost
// Experience

WHERE I'VE WORKED

SNet LabsNoida Jun - Jul 2025
AI & ML Intern
  • Researched and benchmarked foundation models (RTMDetr, DINOv2, SoundStorm)
  • Delivered 7 vision plug-ins: Drowsiness, Sign-Language, Pothole, ANPR, Queue, Flame/Smoke, Smart-Fence
  • Optimized training and inference via quantization and ONNX/OpenVINO
  • Streamlined dataset QA with CVAT for higher model reliability
// Projects

THINGS I'VE BUILT

LLM-based Interactive Interview System

Aug 2025 - Sep 2025

Automated pipeline to parse job descriptions and resumes, generate competency-based interview questions with adaptive follow-ups. Strict JSON grading with score, band, justification, and follow-up question storage. Integrity checks with eye-tracking (OpenCV) and AI-plagiarism detection.

Reduced interview preparation time by 70% and improved candidate evaluation consistency

PythonLangChainOpenAI APIOpenCV
// Projects

THINGS I'VE BUILT

DINOv3 MRI Brain Tumor Diagnosis

Jul 2025 - Aug 2025

Domain-adaptive Vision Transformer for brain tumor classification and segmentation. Achieved 98% test accuracy for 4-class tumor diagnosis. Dice Score of 0.87 and IoU of 0.78 for tumor segmentation masks. Built end-to-end pipeline from data prep to evaluation.

Research paper under review at IEEE. Cross-dataset validation on BRISC-2025 and BraTS

PyTorchHuggingFaceOpenCVDINOv3
// Projects

THINGS I'VE BUILT

Multimodal Cloud-Native RAG Chatbot

Jun 2025 - Jul 2025

Retrieval-augmented assistant with multi-modal support for text, images (BLIP), and voice (Whisper STT + TTS). Built with GROQ Llama-3.3-70B for fast inference. Cloud-native architecture for scalability.

Deployed for internal team use with 95% user satisfaction. Sub-1s response time across all modalities

LangChainGROQ Llama-3.3StreamlitBLIPWhisper
// Skills

MY TOOLKIT

Languages

PythonJavaScriptTypeScriptSQLBash

ML / AI

PyTorchTensorFlowLangChainRAGGPT-4ClaudeLlama-3

Vision

OpenCVDINOv2RTMDetrSegmentation

Backend

Node.jsNext.jsFastAPIRESTWebRTC

MLOps

ONNXOpenVINODockerEdge DeployCI/CD

Tools

GitGitHub ActionsFigmaLinux
// Education & Research

ACADEMIC

B.Tech Computer Science (AI & ML)

UPES Dehradun | CGPA 8.0 | 2022 - Present

DINOv3 Brain Tumor Analysis

IEEE Paper Under Review

Domain-adaptive Vision Transformer with LoRA for classification and segmentation. Cross-dataset validation on BRISC-2025 and BraTS.